Objectives: To establish baseline error rates due to misinterpretation and to identify scenarios in which major errors were most common and potentially preventable.
Methods: Our database was queried over a 3-year period for major discrepancies due to misinterpretation. These were stratified by histomorphologic setting, service, availability/type of prior material, and years of experience and subspecialization of the interpreting pathologist.
Results: The overall discordance rate between frozen section (FS) and final diagnoses was 2.9% (199/6,910). Seventy-two errors were due to interpretation, of which 34 (47.2%) were major. Major error rates were highest on the gastrointestinal and thoracic services. Of major discrepancies, 82.4% were rendered in subdisciplines outside those of the FS pathologist. Pathologists with fewer than 10 years' experience made more errors than those with more experience (55.9% vs 23.5%, P = .006). Major error rates were greater for cases without previous material compared to those with a prior glass slide (47.1% vs 17.6%, P = .009). Common histomorphologic scenarios in which disagreements were made involved discriminating mesothelial cells from carcinoma (20.6%) and accurately recognizing squamous carcinoma/severe dysplasia (17.6%).
Conclusions: To improve performance and decrease future misdiagnoses, monitoring discordances should be a continuous component of surgical pathology quality assurance programs.
Keywords: Accuracy; Discordance; Discrepancy; Error; Frozen section; Intraoperative consultation; Major; Misinterpretation.
© The Author(s) 2023. Published by Oxford University Press on behalf of American Society for Clinical Pathology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.